The Greenland White-fronted Goose Anser albifrons flavirostris

number of birds less the hunting kill. Using only
the initial maximum population count for the year
1967/68, and the mean apparent annual adult
survival rate R 0 (derived from equation 3 above),
the population size for each successive year t+1
was calculated as follows:
68
N t+1 = (R 0 N t (1+J t+1 )) – K t+1 (4)
where K t+1 represents the hunting kill in year t+1,
since maximum numbers usually occur at Wexford
in mid winter after the finish of the hunt
(MS9). We defined K t+1 as the number of geese
recorded shot at Wexford Slobs and Harbour each
winter plus 20% (see above). The model was then
used to generate changes in population size given
observed values for J t+1 up to 1999.
Substituting a constant annual apparent adult
survival rate of 0.884 throughout in the simple
deterministic model produced a remarkably good
fit to the data until 1990 (Figure 8.2). Although
apparently overestimating population size slightly
in the early 1970s, the model shows good correspondence
to the observed values during the period
before and after hunting was banned at Wexford.
After 1990, the model no longer describes
the population development. As there has been
no resumption of hunting in Ireland, this is not
linked to shooting mortality on the wintering
grounds. Rather, it is known from individual
marking that 75% of the 1989 cohort of marked
juveniles failed to return to the wintering grounds
the following season, and adult mortality was also
Maximum annual winter count at Wexford
14000
12000
10000
8000
6000
4000
2000
0
maximum counts
model 0.884 add
model 0.884 add +1989 mortality
1965 1970 1975 1980 1985 1990 1995 2000
Figure 8.2. Model of changes in abundance of Greenland
White-fronted Geese wintering at Wexford Slobs,
based on constant annual adult rate of 0.884 ( ), compared
with the actual annual census counts for the
years 1967/68-1998/99 ( ). The outputs from the
model incorporating the low survival of birds known
from losses of collared individuals following the winter
of 1988/89 are shown as . See text for full details.
10% higher that year (see next section). If the collared
birds were representative of the wintering
numbers at Wexford as a whole, this would have
resulted in some 1900 fewer birds returning in
winter 1990/91. If this exceptional loss is added
in to the model, the fit is greatly improved (see
Figure 8.2).
This approach is extremely simplistic and no attempt
has been made to test goodness of fit of
this model against alternative models. Nevertheless,
the results underline the sensitivity of such
long lived birds to small changes in annual survival/return
rates. The simple model using a constant
annual adult return rate (88.4% percent per
annum) and the assumption of completely additive
hunting mortality to the Wexford wintering
site described the population changes extremely
well in the period prior to and immediately following
the cessation of hunting at this site. The
anomalous deviation in very recent years appears
to be at least partly explained by a year of very
low survival, especially amongst juveniles, following
the summer of 1990, known from population
trends elsewhere and from lowered survival
rates of neck-collared birds. There is thus
reasonable evidence to suggest that hunting at
Wexford may have been additive for the years
1970 to protection on the wintering grounds in
1982, and that the levels of kill experienced in that
period resulted in no clear trend in numbers
(MS14). However, with the cessation of hunting,
Wexford numbers immediately increased, consistent
with expectations if hunting mortality was
additive at this site. Very recent stabilisation and
slight declines in the numbers of geese wintering
at Wexford are consistent with a high mortality
event after winter 1989/90 and with observed
declines in fecundity there (see chapter 6).
8.4 Current measures of annual survival
rates based on individual histories
The first attempt to measure adult annual survival
rates in Greenland White-fronted Geese was
by Boyd (1958), using the Haldane (1955) method
(as the numbers of geese ringed were not known).
He estimated annual adult survival to be 66.1%
(± 3.6 SE) based upon ringing recoveries of birds
ringed and recovered (based largely on hunting
returns) during 1946-1950. Subsequent analysis
using similar methods for all recoveries from
1946-1974 found survival rates of 76.7% (± 3.4 SE,
MS6). Both of these estimates were generated